Recently we played with AI algorithms to see how we can leverage them to the use of our CCM.
One solution we came up with is to be able to predict what could be the next best action during a sales process.
In order to close a deal, you have to go through multiple steps that could be: sending an email, having a phone call, meeting, chat, etc. These combinations of interactions are what will lead you to close a deal.
The more you close deals, the more you end up with creating a pattern. So we decided to pull the data of our own CCM and feed it to the AI algorithm for it to digest and give us what could be the next best action.
In order to make the prediction as accurate as possible we took only deals that are won and the result was promising.
We will, of course, need to improve the solution by taking into account more parameters like the customers preferred canal of communication, the habit of the Sales Rep, etc.
How it works
The extracted data is a list of series, represented by a letter (C, E, M, C) stands for Call, then Email, then Meeting and then finally Call.
So we took the list of series related to all our deals that are won and trained the AI algorithm.
After a while, we took the series of an ongoing deal and made a request to the AI algorithm to predict the next best "letter", which represented our next best action.
What’s interesting here is that the more we use the algorithm the more accurate it will be, based on the data we collect.
This feature will be integrated into our CCM platform very soon, stay tuned!
The next step will be to use the same concept for doing some upsells, by offering products that are bought by other customers, etc.